nicolehhy / SQL

SQL Queries used during internship

Background

When I was a Data Analyst Intern in InKe(a Chinese mobile app allowing users to watch live video broadcast on smartphones), I was mainly in charge of data pulls, data cleaning and building dashboards. I was also responsible for developing various KPIs of the key businees performance indicators that everyone else can use to taclke businees problems.

The following are parts of the takses I touched during the internship. Most of them needed to understand what problem it is, what kind of data I need, how to get them, how to make sure the data integraty and develope some valuable insights through data visulization using Tableau thus to communicate with other departments on some projects more effectively.

The analysis needs that I was given to help different departments tackle business problems

• Feed Strategy(new feature from product department)

  • project introduction
    This folder shows the process of capturing data related to feed strategies which had been released for 2 weeks.
  • What I needed to analyze are the followings:
  1. Number of people who post feeds, Number of comments, and Number of likes per day
  2. Average number of posts per day for different publisher levels
  3. Average number of posts per day for different city levels
  4. The conversion of posting feeds next week from the publishers who posted feeds this week
  5. Publishers'Preferences for Publishing Contents in Different Cities

  The purpose was to analyze Whether the new online feed strategy can help users retain, and whether it can quickly establish a further network between users.

• Posts in App(UCG encouraging strategy from user growth team)

  • project introduction
    When users post feeds, they used the strategy of leaving comments by robots to encourage users to continue to use this feed feature, taking advantage of users' nature in loving being noticed.
  • What I needed to analyze are the followings:
  1. Compared the retention rate on this feed feature before and after using UCG strategy(A/B test)
  2. Compared the Number of people who post feeds, Number of comments, and Number of likes per day before and after using UCG

• Potential Anchor(Anchor encouraging strategy from Strategy team)

  • project introductionv
    The strategy was to increase the number of exposure times of new anchors with high quality live broadcasting on the recommended home page, so as to mine new anchors with high quality and explore more users/followers.
  • What I need to analyze are the followings:
  1. Exposure times, live broadcasting times, live broadcasting duration, followers growth rate per week
  2. Followers conversion rate of new anchors, number of followers giving gifts, amount of gifts

• Push Strategy(Recall loss of users from Marketing team)

  • project introduction
    The project was to recall lost users by sending different text messages(with link) to users who had not logged on to apps for more than a month. There were many strategies in this project. We mainly wanted to analyze which strategy had the best recall effect.
  • What I needed to analyze are the followings:
  1. Delivery rate, arrival rate and click-through rate under different channels, different mobile phone brands and different strategies
  2. Delivery rate, arrival rate and click-through rate for different strategies only
  3. How many people who clicked link in the messages re-login to app, under different strategies

• Recommendation in the same city(User Growth strategy from product department)

  • project introduction
    This Project was to analyze the number of live broadcasts, the length of live broadcasts, and the interaction of fans from the anchors who were in the same city recommendation list, so as to understand the strengths and weaknesses of the strategy
  • What I needed to analyze are the followings:
  1. The number of times users in different cities stay on the same city recommendation pages, the number of videos users watch, the length of time users watch, the number of gifts users give, and the amount of gifts users give.
  2. The time duration of live broadcasting, the average times of live broadcasting, the number of new followers, the number of gifts received, the amount of gifts from the anchors on the the same city recommendation pages.

• The King campaign(Campaign from Operations department)

  • Campaign introduction
    Operations Department invited five famous anchors to participate in a game guessing and lottery draw with fans on one weekend evening. The purpose of this campaign was to encourage users to log in the application again to watch live broadcasting, attract new and old users to participate in the lottery events, thus to increase the time for users to watch live broadcasting, and then analyzed retention rate of users in the next week.
  • What I needed to analyze are the followings:
  1. Number of users participating in the event, total viewing time of users, average viewing time of users, total number of gifts given by users, average number of gifts given by users
  2. The total number of gifts, comments received by anchors and new followers anchors had